Search

virtually, all decisions in financial markets are made in the presence of uncertainty. Stochastic dominance is a well-known concept that is broadly implemented for decision making under uncertainty. Thanks to the advantages of this approach, a decision maker is able to exploit available information related to uncertainties with stochastic nature. In this paper, to cope with the uncertainties in the procedure of decision making for forward contracts in an electricity market, a modified stochastic dominance approach is proposed. In the suggested framework, instead of considering a single value for desirable goal in each scenario of benchmark, a profit interval has been allocated to reflect the economic targets of decision maker. In the next step, the allocated profit interval is utilized for decision making problem in the stochastic dominance framework. In order to find the optimal profit profile in such a setting, a two level optimization structure is suggested. To this end, at the lower level of optimization, three different methods namely the prospect stochastic dominance, conditional value at risk (CvaR) and hypothesis testing are applied to find the optimal profit profiles for a number of benchmarks, those are stochastically generated in an interval around a pre-specified profit benchmark. In the upper level, the optimal profit profile is computed by using the Mean-CvaR method. To show the practical aspects and generalizability of our proposed approach, the methods are applied to two different cases, including a retailer and an electricity producer’s decision making problems to determine their involvement in the futures market by maximizing their expected profit over a given planning horizon, while controlling the risk of profit volatility in the electricity market, raised from the uncertainties in the spot market price and the consumer demand. The performance of suggested framework is evaluated through simulation results and relevant conclusions are drawn. [ABSTRACT FROM AUTHOR]

In an incomplete information game, a big challenge is to find the best way of exploiting available information for optimal decision making of the agents. In this paper, two decision making methods, namely model-based and learning-based bidding strategies, are proposed and compared, for repeated Cournot competition of the generators in a day-ahead electricity market. The sum of the rivals’ offered quantities (SROQ) is considered as the state of the agent and its value is estimated using an adaptive expectation method. In the model-based approach, the convergence of the agents’ strategies to the Nash equilibrium point is also studied in two different cases. In the learning-based approach, the optimal bidding strategy is learned through combination of state estimation and a reinforcement learning method. Using the estimated state (SROQ), the optimal decision is learned through a fuzzy Q-learning algorithm. Through a case study, which is performed on the three-bus benchmark Cournot model, the convergence of the generators’ bids to the Nash-Cournot equilibrium is examined. [ABSTRACT FROM AUTHOR]

The aim of this article is to propose that an intelligent agent can be able to decide properly in an incomplete information repeated Cournot game. The market model and the competitors’ decision models are not known to the players. The proposed agent employs a combination of the k-nearest neighbor (KNN) method and the Bayes classifier to predict the next action of its rivals, using the market decision history. The agent takes the predicted actions as an estimate of its next state and learns the expected payoff of its state-action pairs interactively using the reinforcement learning (RL) algorithm. The results of the proposed agent's competition with two benchmark competitors in different simulated Cournot games are presented. The simulation results show that the proposed agent can significantly earn more payoffs in comparison with the two benchmark agents. [ABSTRACT FROM PUBLISHER]

Abstract: In this paper, an agent-based system for bilateral contracts of energy is proposed. The generating companies submit their offers to the demand companies. The demand companies also submit their bids to the generators. Each load or generator’s agent wants to match with an opponent, which offers the most valuable proposal. However, the problem of simultaneous decision-making causes decision conflicts among the agents. To overcome this conflict, we assume loads as the leaders and generators as the followers. We use Stackelberg game to match the seller and buyer agents. The negotiation process between a buyer and its potential seller will determine the power price between them. This process is carried out through a proposed combined time-behavioral protocol (TBP). With negligible changes in around the agreed price, this protocol can reduce the negotiation time considerably. After successful negotiation, the seller and buyer agents could sign a bilateral contract of energy if the market conditions allow it. The applicability of the proposed method is illustrated through a case study. [Copyright &y& Elsevier]

In a competitive electricity market, suppliers seek the optimal bidding strategy, maximising their individual profit. Due to the uncertainties, unknown parameters and dynamics of electricity market, the optimal bidding strategies cannot be attained through straightforward optimisation methods. In this study, the supply function equilibrium model is considered together with uniform price market clearing mechanism. The interaction among suppliers is modelled as an incomplete information game problem and multi‐agent reinforcement learning (MARL) is utilised to find the optimal bidding strategy of the suppliers in a non‐cooperative game. The proposed framework is an extension, from single agent Markov decision process to Markov game process, which is suitable for studying multi‐agent decision making problems in stochastic environments. In this study, a Markov game model of the electricity market is proposed and a new state‐action and Markov model of electricity market is proposed to built up the MARL environment. In addition, a learning‐based non‐cooperative MARL method is utilised for learning the optimal bidding strategies of the suppliers in a day‐ahead electricity market. The proposed method is successfully applied to IEEE‐30‐bus power system. To examine the statistical significance of the results, the T‐test method is applied to compare the competitive behaviour of the players in both multi‐ and single‐agent frameworks. [ABSTRACT FROM AUTHOR]